Digital Watermarking is a technique utilized to protect digital media from unauthorized use or illegal copying, such as with copyrighted material, for example. Watermarking of digital media may fall into two categories: visible or invisible. Visible watermarks are typically added to digital images to indicate ownership and to thwart unauthorized use of the images. The watermark may comprise the identity of the owner and/or a copyright symbol and date, for example. This type of watermark may be considered a spatial watermark in that the data is embedded spatially in an image, and the watermark signal is distinct from the original image data. Spatial watermarks may not be robust against attacks due to the ability of filtering, removing and/or cropping the data.
Invisible watermarks do not change the image to a perceptible extent. This may be accomplished by minor changes in the least significant bits of the original data. Watermarks that are unknown to the end user may be considered steganographic.
A watermarking process may embed the data in the frequency domain, making it more robust against attack. The technique is somewhat similar to spread spectrum encoding in communications, where the data to embedded may be spread over a multitude of frequencies by modulating the watermark signal with pseudo-noise before adding it to the original data. The low signal amplitude, due to the watermark being invisible, the large bandwidth of the original data (image or video, for example), and the shortness of the watermark message, are all factors that indicate spread spectrum encoding is a logical choice.
Further limitations and disadvantages of conventional and traditional approaches will become apparent to one of skill in the art, through comparison of such systems with the present invention as set forth in the remainder of the present application with reference to the drawings.